A methodology, based on a language's properties, for the selection and validation of a suite of software metrics.

Abstract

Software Engineering has attempted to improve the software development process for over two decades. A primary attempt at this process lies in the arena of measurement. "You can't control what you can't measure" [DEMT82]. This thesis attempts to measure the development of multimedia products. Multimedia languages seem to be the trend of future languages. Problem areas such as Education, Instruction, Training, and Information Systems require that various media allow the achievement of such goals. The first step in this measurement process is the placement of multimedia languages, namely Authorware, in the existing taxonomy of language paradigms. The next step involves the measurement of various distinguishing properties of the language. Finally, the measurement process is selected and evaluated. This evaluation gives insight as to the next step in establishing the goal of control, through measurement, of the multimedia software development process.